ZSM-5 Catalyst Regeneration: Coke Removal Temperature Optimization for MTO, FCC, and Xylene Isomerization

Why Regeneration Temperature Is the Most Important Lever in ZSM-5 Catalyst Economics

ZSM-5 is the workhorse acid catalyst for methanol-to-olefins (MTO), fluid catalytic cracking (FCC) propylene-boost additive, and xylene isomerization. In each application the catalyst spends its life between two states: hot hydrocarbon atmosphere building up carbonaceous deposits (coke), and a hot oxidative atmosphere burning those deposits off. The cost of the second state, the regeneration, is the silent line item in your operating budget. For a 1.83 Mt/y MTO plant with 600 MT of ZSM-5 inventory, regeneration energy alone runs 8 to 18 million USD per year, and the catalyst replacement cost (which is largely driven by how well you regenerate) adds another 12 to 25 million USD per year. Get the regeneration temperature window right and you can push single-charge catalyst life from 100 cycles to 400+ cycles, which cuts catalyst replacement cost by 60%.

ZSM-5 regeneration is harder than FCC USY regeneration or silica gel regeneration for three reasons. First, ZSM-5 has a 10-membered-ring pore system (5.3 x 5.6 Angstrom and 5.1 x 5.5 Angstrom) that traps bulky polyaromatic coke inside the pore intersections. The coke cannot diffuse out without being burned in place. Second, the MFI framework is more sensitive to high-temperature steam than larger-pore zeolites. The framework Al-O bond hydrolyzes above 560 degrees C in the presence of steam, and once an Al leaves the framework it cannot easily go back. Third, ZSM-5 catalyst life in MTO is so short (single cycle 8 to 48 hours) that the regenerator runs every day, hundreds of times per year, and cumulative damage matters more than single-cycle performance.

This article is the operating engineer's reference. We will cover the chemistry of coke on ZSM-5, the 5-stage burnoff protocol we use at Aluminaworld's reference Zibo plant, the kinetics of acid site loss (dealumination), the trade-offs between steam dilution and O2 partial pressure, three industrial case studies (DMTO fluidized bed, FCC additive, xylene isomerization fixed bed), and the QC tests that confirm regeneration is complete without collateral damage. All numbers in this article are typical industrial values; your exact numbers will depend on feed composition, reactor configuration, and coke loading profile. If you want to discuss your specific case, our technical team is one WhatsApp message away.

What Coke on ZSM-5 Actually Is: Soft Coke, Hard Coke, and the Grey Zone

ZSM-5 coke is not a single compound. It is a continuously evolving mixture of polyaromatic and alkyl-aromatic species whose composition shifts with time on stream, temperature, and feedstock. The two operational categories that matter for regeneration engineering are 'soft coke' (low-temperature coke, H/C ratio 0.5 to 0.7) and 'hard coke' (high-temperature coke, H/C ratio 0.2 to 0.4), with a grey zone of 'intermediate coke' (H/C 0.4 to 0.5) in between.

Soft coke is the early deposit. It builds up in the first 5 to 30 minutes of MTO service and is dominated by alkylated benzenes and polymethylbenzenes that form via cyclization and hydrogen transfer in the MFI channel intersections. At MTO conditions (400 to 480 degrees C, 1 to 3 bar, methanol WHSV 1 to 5 h-1), soft coke is 60 to 80% of total coke after 8 hours on stream. The H/C ratio is high (0.55 to 0.70), the deposit is hydrogen-rich, and the burnoff temperature window is 200 to 400 degrees C. Soft coke is the right target for partial regeneration when you want to extend MTO cycle length without paying the full thermal cost.

Hard coke is the late deposit and the one that dominates after long time on stream (50+ hours in MTO, weeks in xylene isomerization). It is essentially graphite-like polyaromatic material with H/C below 0.4, formed by sequential dehydrogenation and condensation of soft coke precursors. Hard coke sits both inside the pores (the small fraction that fits) and on the external crystal surface (the larger fraction that cannot fit). The burnoff temperature window is 450 to 600 degrees C, and complete removal requires either longer time at 540 degrees C or higher temperature up to 580 degrees C. Hard coke is what drives the cumulative dealumination problem because removing it requires holding the catalyst at the top of the safe temperature window.

Intermediate coke (H/C 0.4 to 0.5) is the grey zone that dominates at moderate time on stream (10 to 30 hours) and at intermediate temperatures (350 to 450 degrees C). It is a mix of partially dehydrogenated alkylaromatics and is responsible for the slow activity decline between the soft-coke plateau and the hard-coke plateau. Industrial regenerators that target 'complete' regeneration at 540 to 560 degrees C will burn intermediate coke alongside hard coke, but the energy required and the dealumination risk are higher than for soft-coke-only burn.

The coke composition matters for regeneration because each class burns at a different temperature and with a different exotherm. The heat of combustion for soft coke is approximately 38 to 42 kJ/g (high because of the hydrogen content), while hard coke burns at 28 to 32 kJ/g (low because of the aromatic condensation). The mixed coke on a real MTO catalyst after 24 hours on stream has an effective heat of combustion of 32 to 36 kJ/g. If you do not stage the O2 supply, the bulk burn in the 300 to 400 degrees C range generates a 100 to 200 degrees C adiabatic temperature spike that pushes local catalyst above 700 degrees C and permanently destroys acid sites. The 5-stage protocol we describe later in this article exists to manage exactly this problem.

The Burnoff Temperature Window: 480 to 580 degrees C and Why the Edges Matter

The safe regeneration window for ZSM-5 is narrower than most operators realize. Below 450 degrees C, even extended burnoff (10 to 24 hours) leaves 0.5 to 2 wt% residual carbon on the catalyst because the slow combustion kinetics of hard coke plateau out. Above 600 degrees C, the framework begins to dealuminate at a rate of 0.1 to 0.5 wt% Al per hour in the presence of even 2 vol% steam. The 480 to 580 degrees C window is the operating range, with 520 to 560 degrees C being the most common target for steady-state industrial operation.

Reference: Aluminaworld 5 m3 regeneration pilot, Si/Al 100 ZSM-5, 5 to 12 wt% initial coke loading, 2 vol% steam co-feed. Acid site loss measured by NH3-TPD against fresh reference.

The acid site loss numbers in the table are the single most important column. At 540 to 580 degrees C (the standard operating window), each hour of soak time removes 0.2 to 0.6% of your remaining strong acid sites. If your regeneration cycle is 8 hours at 560 degrees C, that is 1.5 to 5% acid site loss per cycle. Over 200 cycles, that compounds to 60 to 80% loss - which is why we say 200 to 600 cycles is the realistic life range. If you push to 600 to 700 cycles by running at 600 to 620 degrees C to speed up the burn, you double or triple the per-cycle acid site loss and your 200-cycle catalyst is now your 50-cycle catalyst.

Why 600 degrees C is the absolute hard limit

Dealumination kinetics for MFI framework zeolites in the presence of steam follow an Arrhenius relationship with an apparent activation energy of 90 to 130 kJ/mol. Below 550 degrees C the rate is slow enough that it does not matter for industrial cycle life (0.05 to 0.2% acid site loss per hour). Above 600 degrees C the rate doubles every 10 to 15 degrees C. At 650 degrees C, dealumination is 10 to 30 times faster than at 550 degrees C, and the framework can lose 30% of its acid sites in 30 minutes. This is why experienced operators cap regenerator bed temperature at 580 degrees C even though the CO2 in flue gas might not yet be at baseline - they accept 0.05 to 0.2 wt% residual carbon in exchange for not destroying 10% of their acid sites.

The other problem above 600 degrees C is metal migration. If your ZSM-5 has been impregnated with Cu, Fe, Zn, or Pt (for SCR, xylene isomerization, or specialty applications), these metal species become mobile above 600 degrees C and migrate from the ion-exchange sites to form metal oxide clusters on the external surface. Cu-ZSM-5 (SCR catalyst) is the most sensitive - Cu migrates above 580 degrees C and forms CuO clusters that block pores. Pt-ZSM-5 sinters above 620 degrees C with a 30 to 50% loss in Pt dispersion. Zn-ZSM-5 (used in some xylene isomerization recipes) loses 60 to 80% of its Zn above 650 degrees C. The 'safe' temperature for metal-loaded ZSM-5 is therefore 540 to 560 degrees C, not 580 degrees C.

The 5-Stage Burnoff Protocol That Recovers 92 to 96% of Acid Sites

Industrial ZSM-5 regeneration is a sequence of five stages, each with its own temperature, gas composition, and duration. The protocol is staged to manage the exotherm and minimize dealumination while still achieving complete coke removal. Below is the protocol we use for MTO and xylene isomerization service at Aluminaworld's reference plant, with deviations noted for FCC additive service at the end.

Stage 1: Drying and preheat (ambient to 200 degrees C, 1.5 to 2 hours)

Heat the catalyst bed from ambient to 200 degrees C at 1 to 3 degrees C per minute in dry N2 (or flue gas recycle) at GHSV 200 to 500 h-1. The purpose is to remove physisorbed water and light hydrocarbons. The temperature must stay below 200 degrees C because any residual methanol, DME, or hydrocarbon vapor will ignite in the presence of even trace O2 if the bed is above 200 degrees C. The exit gas should be monitored for hydrocarbon slip; once it drops below 100 ppm, proceed to Stage 2. Typical water evolved: 2 to 5 wt% of catalyst weight. Typical duration: 1.5 to 2 hours.

Stage 2: Soft coke burn (200 to 400 degrees C, 2 to 4 hours)

Introduce 0.5 to 1.0 vol% O2 in N2 at 200 degrees C. Hold for 30 minutes to confirm no exotherm. Ramp at 0.5 to 1.0 degrees C per minute to 400 degrees C. The soft coke begins burning at 220 to 250 degrees C, with peak exotherm at 320 to 380 degrees C. The flue gas CO2 will rise to 2 to 5 vol% during the peak and then fall. Hold at 400 degrees C until CO2 drops below 0.3 vol%. Total Stage 2 time is 2 to 4 hours. This stage removes 40 to 60% of the total coke. Acid site loss is negligible (< 0.1%) because the temperature is well below the dealumination onset.

Stage 3: Hard coke burn (400 to 540 degrees C, 2 to 4 hours)

Increase O2 to 1.5 to 2.5 vol% and ramp at 0.5 to 1.0 degrees C per minute to 540 degrees C. The intermediate coke burns at 400 to 460 degrees C, hard coke at 460 to 540 degrees C. The exotherm is smaller than Stage 2 because hard coke has a lower heat of combustion. Flue gas CO2 will peak at 1 to 2 vol% and then fall. Hold at 540 degrees C until CO2 drops below 0.2 vol%. Stage 3 removes another 35 to 50% of total coke (cumulative 90 to 95%). Acid site loss in this stage is 0.1 to 0.3% per hour, mostly from the upper end of the temperature range.

Stage 4: Final burnoff (540 to 580 degrees C, 1 to 2 hours)

Switch to 5 to 10 vol% O2 in N2 (or 21 vol% O2 / air if the bed is small and the heat removal is adequate). Hold at 540 to 580 degrees C for 1 to 2 hours. The remaining 5 to 10% of coke (the most graphitic fraction, H/C 0.2 to 0.3) burns here. The exotherm is small (0.2 to 0.5 degrees C temperature rise) because so little coke is left. Acid site loss accelerates in this stage (0.3 to 0.6% per hour) but the duration is short. Stop when flue gas CO2 is below 0.05 vol% (below 500 ppm) and CO is below 50 ppm. Residual carbon on the catalyst should now be below 0.10 wt%.

Stage 5: Cool down and re-equilibrate (580 degrees C to ambient, 2 to 3 hours)

Switch to dry N2 (or dry air if the catalyst can tolerate brief O2 exposure at lower temperature). Cool at 1 to 2 degrees C per minute to ambient. Do not pass air over the hot catalyst during cooldown because the heat of O2 chemisorption on partially reduced metal sites can spike temperature. Below 100 degrees C, the catalyst can be exposed to ambient air. The cool-down rate matters less than the heat-up rate, but rapid cooldown (quench) should be avoided because the thermal stress on the binder (alumina, attapulgite, kaolin) can crack the extrudate.

Total cycle time for the 5-stage protocol is 8 to 14 hours depending on initial coke loading. The protocol recovers 92 to 96% of the strong acid sites on the first regeneration and 88 to 94% on subsequent regenerations (the cumulative dealumination is the difference). Compare this to a linear 2 degrees C per minute ramp from ambient to 580 degrees C in 4 vol% O2, which recovers only 75 to 88% of acid sites and produces visible hot spots in the 250 to 400 degrees C region of the bed. The staged protocol trades cycle time for acid site recovery, and the trade-off pays for itself many times over in extended catalyst life.

Oxygen Partial Pressure: Why 0.5 to 4 vol% Beats 21 vol% Air

The temptation in regeneration is to use air (21 vol% O2) to maximize burn rate. The reality is that air-regeneration is appropriate for FCC fluidized bed regenerators (where the catalyst inventory is small, the heat removal is excellent, and the catalyst residence time in the regenerator is only 1 to 5 minutes) but is wrong for fixed-bed MTO swing reactors and for any application where the catalyst stays in the regenerator for hours. The reason is heat management: the adiabatic temperature rise from burning 10 wt% coke in air is 1100 to 1300 degrees C - well above the dealumination onset, the framework collapse temperature, and the metal migration temperature. Even at 4 vol% O2 in N2, the adiabatic temperature rise is 250 to 400 degrees C, which is still too much for a single stage but is manageable with the staged O2 ramp described in Section 4.

The right O2 partial pressure depends on the stage of regeneration. For Stages 2 and 3 (bulk burn), 0.5 to 2.5 vol% O2 in N2 keeps the local exotherm below 80 to 150 degrees C. For Stage 4 (final burnoff), 5 to 10 vol% O2 is acceptable because so little coke is left that the exotherm is small. For Stage 5 (cooldown), 21 vol% O2 / air is fine because there is no combustion. The economics matter too: regenerating in air means paying to heat 79 vol% N2 to 580 degrees C and losing it to the stack, while a 4 vol% O2 / 96 vol% N2 mixture (once the coke is gone) is mostly recirculated, cutting regeneration gas cost by 40 to 60%.

For large DMTO fluidized bed regenerators (the dominant industrial MTO configuration in China), the O2 profile is different. The fluidized bed regenerator runs continuously at 580 to 620 degrees C with 2 to 4 vol% O2 in flue gas recycle, plus air injected at the bottom for the coke burn. The catalyst residence time in the regenerator is 1 to 3 minutes, so each catalyst particle is exposed to high temperature for a short time. Cumulative dealumination is controlled by the inventory-to-throughput ratio, not by per-pass exposure. A 1.83 Mt/y DMTO unit with 600 MT catalyst inventory running 1.5 cycles per hour (40 min on stream + 20 min regen) achieves 0.05 to 0.15% acid site loss per cycle, which is the gold standard for industrial regeneration.

Steam Dilution: The Double-Edged Heat Sink and Dealumination Agent

Steam is the most common additive to regeneration gas for two reasons. First, it has a heat capacity roughly twice that of N2 (1.87 kJ/kg-K vs 1.04 kJ/kg-K at 500 degrees C), so adding 20 to 30 vol% steam to the regeneration gas cuts the adiabatic temperature rise from coke combustion by 30 to 40%. Second, steam gasifies soft coke via steam reforming and water-gas shift at 350 to 500 degrees C, which removes the most reactive coke species before they can burn exothermically in the O2 stage. Both effects are real and both are used industrially.

The dark side of steam is dealumination. The MFI framework hydrolyzes in steam above 500 degrees C, with the rate doubling every 10 to 15 degrees C. Every 1 vol% of steam at 560 degrees C removes roughly 0.05 to 0.15% of framework Al per hour. The mechanism is well understood: H2O attacks the framework Al-O bond, forming Al-OH and Si-OH terminal groups that are not re-connected on cooling. The result is an extra-framework Al (EFAl) species that may be re-incorporated at lower temperature but is permanently lost above 600 degrees C.

The optimal steam strategy is therefore: (1) Add 15 to 30 vol% steam to Stage 2 (200 to 400 degrees C) to gasify soft coke and limit exotherm. (2) Reduce steam to 5 to 10 vol% in Stage 3 (400 to 540 degrees C) as the dealumination risk increases. (3) Cut steam to 0 to 3 vol% in Stage 4 (540 to 580 degrees C) to avoid cumulative dealumination. (4) Eliminate steam entirely in Stage 5 (cooldown) because no combustion is happening. Industrial regenerators that use 'pure steam' regeneration (90+ vol% steam throughout) look white at the end but have lost 30 to 50% of their acid sites to dealumination. The catalyst comes out clean but useless.

There is a special case where higher steam is acceptable: in situ steam regeneration of xylene isomerization catalyst. Xylene isomerization runs on Pt- or Re-loaded ZSM-5 at 350 to 450 degrees C with hydrogen co-feed, and the coke is much softer (H/C 0.5 to 0.7) than MTO coke. In this application, regeneration in 80 to 100 vol% steam at 450 to 500 degrees C for 4 to 8 hours removes 90 to 95% of the coke with less than 2% acid site loss per cycle. The Pt or Re metal does not migrate because the temperature is well below 580 degrees C. This is one of the few cases where 'pure steam' is the right answer, and it only works because the regeneration temperature is low enough that dealumination is negligible.

Dealumination Kinetics: Why Each Cycle Costs 0.5 to 2.5% of Your Acid Sites

Dealumination is the silent killer of ZSM-5 catalyst life. The phenomenon is straightforward: each regeneration cycle at 540 to 580 degrees C with even 2 to 5 vol% steam in the gas removes 0.5 to 2.5% of the framework Al. After 100 cycles, 30 to 70% of the original strong acid sites are gone. The catalyst is still white, still has the MFI structure by XRD, but its activity and selectivity have drifted to those of a higher-Si/Al material. The operating consequence: propylene selectivity in MTO increases (because higher Si/Al favors propylene over heavier products), but methanol conversion at constant WHSV drops, requiring higher reactor temperature to maintain conversion. Eventually the reactor temperature hits the metallurgical limit of the reactor tubes and the catalyst must be replaced.

The dealumination rate depends on three factors: temperature, steam partial pressure, and time. Temperature is the dominant lever. At 540 degrees C, 2 vol% steam, 4 hours exposure, the dealumination rate is 0.10 to 0.20% acid site loss per hour. At 560 degrees C, same steam and time, it is 0.25 to 0.40%. At 580 degrees C, it is 0.50 to 0.80%. At 600 degrees C, 1.0 to 1.8%. At 620 degrees C, 2.0 to 3.5%. The Arrhenius activation energy for this process in MFI is 95 to 120 kJ/mol, which is high enough that small temperature excursions have outsized effects.

Steam partial pressure matters but is sublinear. Going from 0 to 5 vol% steam roughly doubles the dealumination rate. Going from 5 to 15 vol% adds another 50%. Going from 15 to 50 vol% adds another 30%. The non-linearity comes from the fact that the dealumination reaction produces water as a product, so once some Al-O bonds have broken the local steam partial pressure is already elevated. The implication: there is little point in running steam above 15 to 20 vol% in the bulk-burn stages because the marginal dealumination cost outweighs the marginal exotherm-reduction benefit.

Time matters linearly (per the dealumination kinetics being roughly first-order in time at constant T and steam), but with a saturation effect: after 4 to 6 hours at 580 degrees C in 5% steam, the framework has lost the most vulnerable Al sites, and the rate slows. The implication is that extending Stage 4 burnoff to chase the last 0.05 wt% of residual carbon is usually a bad trade-off. Stop at 0.10 wt% residual C and accept that you did not get the absolute last 50 ppm. The acid sites you save by stopping are worth more than the carbon you leave behind.

What dealumination looks like in QC data

Dealumination is visible in three QC tests. First, ICP-OES bulk Si/Al ratio does not change (because the Al is still in the sample, just not in the framework), but the framework Si/Al can be measured by 29Si MAS-NMR and it does increase. Second, NH3-TPD shows a sharp drop in the strong acid site peak (centered at 350 to 420 degrees C) and a relative increase in the weak acid site peak (centered at 180 to 250 degrees C). The weak acid sites are framework Si-OH groups that form when the neighboring Al is removed. Third, 27Al MAS-NMR shows a growing peak at 0 ppm (octahedral extra-framework Al) and sometimes a small peak at 30 ppm (tetrahedral extra-framework Al). The 0 ppm peak is the dealumination signature.

EFAl is not always bad. In some reactions, EFAl acts as a Lewis acid site that promotes hydrogen transfer and coke precursor formation (which is bad for MTO) or that promotes dehydrocyclization (which is good for some alkylation reactions). The take-away: do not assume dealumination is uniformly negative. The right approach is to measure the impact of EFAl on your specific reaction. For MTO, EFAl is generally harmful because it accelerates coking. For alkylation and some isomerization, it can be neutral or even slightly beneficial. The decision of when to dump catalyst is reaction-specific, not framework-specific.

Coke Loading Profiles by Application: 2 to 18 wt% and Why It Matters

The coke loading on a ZSM-5 catalyst when it goes to regeneration is a strong function of the application and the reactor configuration. The numbers below are typical industrial ranges from Aluminaworld customer data, 2024 to 2026 production.

Coke loading measured by CHN analysis on spent catalyst; cycle length and regen time from industrial customer logs 2024 to 2026.

The coke loading at regeneration has two practical implications. First, higher coke loading means more exotherm in Stage 2, which means the O2 supply must be turned down to avoid hot spots. The rule of thumb is 1 vol% O2 per 5 wt% coke on catalyst. So a 10 wt% coke loading needs 2 vol% O2 max in Stage 2, while a 5 wt% coke loading can handle 1 vol% O2 without exceeding 580 degrees C bed temperature. Second, higher coke loading means longer Stage 3 (hard coke burn) because the absolute amount of hard coke is higher. The total cycle time scales roughly linearly with initial coke loading above 5 wt%.

For DMTO fluidized bed units, the coke loading is kept low (4 to 8 wt%) deliberately. The fluidized bed regenerator design and the continuous catalyst circulation make it more efficient to run shorter cycles at lower coke than to run longer cycles at higher coke. The cost is more frequent regeneration and more catalyst attrition from circulation, but the benefit is gentler per-cycle conditions and longer total catalyst life (1000+ cycles in some DMTO units). The fixed-bed swing MTO design has the opposite trade-off: less frequent regeneration at the cost of higher per-cycle acid site loss. There is no universal right answer; the right choice depends on the rest of the plant design.

Three Industrial Case Studies

Case 1: 1.83 Mt/y DMTO fluidized bed unit (China, 2024)

A 1.83 Mt/y DMTO (methanol-to-olefins, Sinopec DMTO process) unit in Inner Mongolia operates with 600 MT of ZSM-5 catalyst inventory (template-free, Si/Al 180, hierarchical mesoporous, supplied by Aluminaworld). The fluidized bed reactor runs at 480 to 500 degrees C, 1.5 bar, WHSV 2 h-1, with continuous catalyst circulation between reactor and regenerator. The regenerator runs at 580 to 600 degrees C, 2 vol% O2 in flue gas recycle, with air injected at the bottom for the coke burn. Coke loading on the catalyst going to regeneration is 5 to 7 wt%; coming back from regeneration it is below 0.15 wt%.

After 18 months of operation (approximately 12,000 regeneration cycles), the catalyst has lost 28% of its initial strong acid sites by NH3-TPD and the framework Si/Al has drifted from 180 to 215. Methanol conversion at constant WHSV has dropped from 99.8% to 99.2%, and the operators have raised reactor temperature by 8 degrees C to maintain conversion. Propylene selectivity has increased by 1.2 percentage points (from 41.5 to 42.7 wt% of hydrocarbons), which is a small bonus from the effective increase in Si/Al. The plan is to operate the catalyst for another 12 to 18 months before dumping and reloading with fresh ZSM-5, giving a total catalyst life of 30 to 36 months and approximately 20,000 to 24,000 regeneration cycles.

Key learning: the staged O2 supply in the regenerator (2 vol% O2 in flue gas recycle, plus 21 vol% O2 air at the bottom) keeps the bed temperature in the 580 to 600 degrees C window and limits dealumination to 0.10 to 0.15% per cycle. Operators who tried to speed up the regeneration by using 5 to 8 vol% O2 throughout saw 0.4 to 0.6% acid site loss per cycle and had to dump catalyst after 6 to 8 months. The original staged protocol is now the standard across all DMTO units in China.

Case 2: 100,000 MT/y polypropylene plant with FCC additive (Middle East, 2025)

A polypropylene plant in Saudi Arabia uses a FCC catalyst with 2.5 wt% ZSM-5 additive (Aluminaworld template-free, Si/Al 50) for propylene boost. The FCC unit is a conventional riser-regenerator design with catalyst circulation at 1 to 2 wt% per minute. The regenerator runs at 700 to 720 degrees C in air (21 vol% O2), which is much hotter than ZSM-5 standalone regeneration but the catalyst residence time is only 30 to 90 seconds per pass, so the cumulative thermal exposure per cycle is small.

After 8 months of operation, the ZSM-5 additive has been through approximately 35,000 FCC regeneration cycles (each cycle being 30 to 90 seconds in the regenerator). The ZSM-5 has lost 18% of its initial strong acid sites and the propylene boost has dropped from +4.8 wt% to +4.0 wt% (still positive but trending down). The plan is to continue for another 4 to 6 months and then top up the ZSM-5 additive by 30% to restore activity. The full ZSM-5 inventory will be replaced after 18 to 24 months.

Key learning: FCC additive ZSM-5 sees many more regeneration cycles than fixed-bed MTO, but each cycle is much shorter. The cumulative dealumination is 0.001 to 0.003% per cycle (versus 0.10 to 0.15% per cycle in DMTO), so the total life is longer. The trade-off is that the ZSM-5 also sees the FCC feed poisons (V, Ni, Fe) which are not present in MTO service, so the actual life-limiting factor in FCC is metal poisoning rather than dealumination.

Case 3: 1.2 Mt/y paraxylene plant, xylene isomerization fixed bed (India, 2026)

A paraxylene plant in Gujarat operates a xylene isomerization unit with Pt-loaded ZSM-5 (Pt 0.5 wt%, Si/Al 35, supplied by Aluminaworld). The reactor runs at 380 to 420 degrees C, 10 to 20 bar, H2/oil ratio 4, WHSV 3 h-1. Cycle length between regenerations is 6 to 9 months. The regeneration is in situ steam regeneration (no O2, 90+ vol% steam at 470 to 500 degrees C for 8 hours, followed by N2 purge and reactor cooldown).

After 3 years of operation (5 regenerations), the catalyst has lost 12% of its initial strong acid sites and 8% of its Pt dispersion. The paraxylene selectivity has dropped from 88.5 to 86.8 wt% of xylene products, and the operators have raised reactor temperature by 6 degrees C to maintain xylene conversion. The plan is to operate for another 2 to 3 years (3 to 4 more regenerations) before reloading, giving a total catalyst life of 5 to 6 years and 8 to 9 regenerations. The longer cycle length and lower regeneration severity make xylene isomerization the easiest of the three applications to manage from a regeneration standpoint.

Key learning: in situ steam regeneration of Pt-ZSM-5 in xylene isomerization is one of the few cases where pure-steam regeneration is correct. The temperature is low enough (470 to 500 degrees C) that dealumination is negligible (0.2 to 0.5% per regeneration versus 0.5 to 2.5% for O2 regeneration at 540 to 580 degrees C). The trade-off is incomplete coke removal (5 to 10% residual carbon after steam regeneration), but the residual carbon does not affect the xylene isomerization reaction because the reaction is hydrogen-transfer-driven and not very sensitive to micropore blockage.

Total Cost of Ownership: Why Regeneration Optimization Pays for Itself in 3 to 6 Months

The economic case for optimized ZSM-5 regeneration is straightforward. Below is a worked example for a 1.83 Mt/y MTO plant with 600 MT ZSM-5 inventory, comparing three regeneration strategies.

Reference: 1.83 Mt/y MTO unit, 600 MT ZSM-5 inventory, ZSM-5 cost 9 USD/kg delivered. Energy cost 0.07 USD/kWh. Cycles calculated from MTO swing operation (1 cycle per 24 hours).

The optimized 5-stage protocol saves 25 to 60 million USD per year in catalyst replacement cost alone, and another 1 to 6 million USD per year in regeneration energy. The cumulative saving over a 6-year catalyst life is 60 to 180 million USD. The capital cost of switching to the optimized protocol is essentially zero - the same regenerator hardware is used, only the operating procedure changes. The payback period for the procedural change is 1 to 3 months, and the savings continue for the life of the plant.

The most common reason operators give for not adopting the optimized protocol is 'we have always done it this way and the catalyst lasts long enough'. The math does not support this. Even at the optimistic end of the 'standard regen' column, the optimized protocol saves 24 million USD per year, which is more than the cost of a new regenerator. The optimized protocol also produces less NOx and CO2 per regeneration cycle, which has direct compliance value in jurisdictions with emissions caps.

Eight Common Mistakes When Regenerating ZSM-5

1. Ramping too fast through the soft coke burn range (200 to 400 degrees C). The exotherm is highest in this range because soft coke has the highest heat of combustion. A linear 2 to 3 degrees C per minute ramp overshoots the setpoint by 50 to 200 degrees C in this region, and the resulting hot spots destroy 5 to 10% of acid sites per cycle. Use a staged ramp of 0.5 to 1.5 degrees C per minute through this range.
2. Using air (21 vol% O2) for fixed-bed regeneration. Air gives 1100 to 1300 degrees C adiabatic temperature rise for 10 wt% coke loading. There is no fixed-bed regenerator that can dissipate this safely. Use 0.5 to 10 vol% O2 in N2 staged over the protocol.
3. Running 100% steam regeneration at 540 to 600 degrees C. The catalyst comes out white but the dealumination is severe. Pure steam regeneration is only correct for low-temperature applications (<500 degrees C, e.g. xylene isomerization with Pt- or Re-loaded catalyst).
4. Skipping the soft coke burn stage. Some operators jump from 200 degrees C directly to 540 degrees C in 4 vol% O2 to save time. The result is a massive exotherm in the 300 to 450 degrees C range as soft and intermediate coke burn simultaneously, with bed hot spots above 700 degrees C. Total cycle time may be shorter (6 hours vs 12 hours) but acid site loss is 3 to 5x higher.
5. Continuing Stage 4 burnoff until flue gas CO2 is at instrument zero. The last 0.05 to 0.1 wt% of carbon is the most graphitic and the most expensive to remove in terms of dealumination cost. Stop when residual C is below 0.10 wt% and accept the trade-off.
6. Not analyzing the regenerated catalyst before re-loading. The five QC tests (CHN, ICP-OES, BET, NH3-TPD, XRD) take 4 to 6 hours and cost 200 to 500 USD per sample. Skipping them to save time is penny-wise and pound-foolish - one bad batch going back into the reactor can cost 100,000 to 500,000 USD in unscheduled downtime.
7. Confusing brown catalyst (EFAl) with grey catalyst (incomplete burnoff). Brown discoloration after regeneration is usually extra-framework Al (Lewis acid, often catalytic poison) and cannot be removed by further burnoff. Grey discoloration is residual carbon and CAN be removed by further burnoff. If the catalyst is brown, do not re-load it - send it for acid wash and re-exchange, or discard.
8. Not isolating the regenerator from the reactor before starting burnoff. Residual hydrocarbon vapor in the regenerator can ignite when O2 is introduced above 200 degrees C. The 200 degrees C 'no exotherm' hold in Stage 2 exists to confirm that all hydrocarbons have been desorbed before O2 burnoff starts. Skipping this hold is the most common cause of regenerator fires.

Quality Control: Five Tests on Every Regenerated Batch

The five QC tests on every regenerated ZSM-5 batch are listed in the order they should be run, with the cost and turnaround time for each. Most of the tests can be run on a 5 to 10 g sample taken from the regenerated catalyst bed after cooldown.

1. CHN analysis for residual carbon (USD 30 to 80 per sample, 2 to 4 hours). Combust the sample in pure O2 at 1100 degrees C, measure CO2 by IR detector. Target: below 0.10 wt% C. A value above 0.5 wt% means the catalyst needs more burnoff time. A value above 2 wt% means the regeneration was a failure and the catalyst should not be re-loaded.
2. ICP-OES for Si/Al ratio and trace metal (USD 50 to 150 per sample, 4 to 8 hours). Acid-digest in HF/HNO3/HCl, analyze by ICP-OES. Si at 251.6 nm, Al at 396.2 nm. Target: Si/Al within 5% of fresh (i.e. if fresh is 100, regenerated should be 95 to 105). A value above 10% drift means significant dealumination. Also report Na, Fe, Ca, Mg, V, Ni - these are feed poisons and should not increase cycle over cycle.
3. BET surface area and micropore volume (USD 100 to 200 per sample, 6 to 12 hours). N2 physisorption at 77 K per ISO 9277. Target: BET within 8% of fresh (e.g. if fresh is 400 m2/g, regenerated should be 368 to 432 m2/g). Micropore volume by t-plot method, target within 10% of fresh. Loss above 15% indicates framework collapse (over-calcination).
4. NH3-TPD for acid site distribution (USD 150 to 300 per sample, 8 to 16 hours). Heat sample from 100 to 600 degrees C in NH3 flow, monitor desorption with TCD or MS. Target: total acid site count within 15% of fresh, strong acid site peak (350 to 420 degrees C) within 20% of fresh, weak acid site peak (180 to 250 degrees C) within 25% of fresh. A sharp drop in strong acid sites is the dealumination signature. A growing weak acid peak is also dealumination (Si-OH formation).
5. XRD for relative crystallinity (USD 50 to 150 per sample, 2 to 4 hours). Diffractometer scan from 5 to 50 degrees 2-theta, Cu K-alpha, compare peak intensities at 7.9, 8.9, 23.1, 23.9, 24.4 degrees against in-house reference. Target: above 92% for template-free ZSM-5, above 94% for template-driven ZSM-5. A drop below 88% indicates severe thermal damage and the catalyst should not be re-loaded.

Optional but recommended: 27Al MAS-NMR (USD 300 to 800 per sample, 1 to 3 days) for framework vs extra-framework Al quantification. Target: above 80% tetrahedral framework Al. A value below 70% indicates the catalyst is approaching end-of-life even if the other QC tests are within spec.

Decision rules after QC

Once the QC results are in, the decision is one of three: re-load the catalyst as-is, regenerate it again, or dump it. The rules below are based on Aluminaworld's customer data from 2024 to 2026.

• Re-load as-is: All five QC tests within spec, residual C below 0.10 wt%, no visual discoloration.
• Re-regenerate: Residual C 0.10 to 0.50 wt%, all other tests within spec. Send back to Stage 4 of the protocol for another 2 to 4 hours of burnoff.
• Acid wash and re-exchange: Brown discoloration, BET loss above 15%, NH3-TPD weak acid peak growing. Wash in 0.5 M HNO3 at 60 degrees C for 1 hour, then re-exchange with 1 M NH4NO3 at 80 degrees C for 2 hours, then recalcine. This recovers 50 to 70% of the lost strong acid sites in some cases.
• Dump: Residual C above 2 wt%, BET loss above 25%, NH3-TPD strong acid peak loss above 50%, XRD crystallinity below 85%. The catalyst is at end of life and should be replaced. Spent ZSM-5 can be sent for Al recovery (acid leach + Al2O3 precipitation) or used as a low-grade binder in cement or brick.

Standards and Reference Methods

The relevant international standards for ZSM-5 regeneration and quality testing include:

• ASTM D7582-15 — Standard test methods for proximate analysis of coal and coke by thermogravimetry. For residual carbon determination on regenerated catalyst.
• ISO 9277:2022 — Determination of specific surface area of solids by gas adsorption (BET method). For BET surface area measurement.
• ISO 13320:2020 — Particle size analysis by laser diffraction. For PSD measurement of spent and regenerated catalyst.
• ISO 17294-2:2016 — Application of inductively coupled plasma mass spectrometry (ICP-MS). For trace metal analysis (Fe, V, Ni, Na, Ca, Mg).
• ASTM UOP 874-13 — UOP method for zeolite crystallinity by XRD. Standard reference for MFI crystallinity.
• GB/T 30470-2013 — Chinese national standard for ZSM-5 zeolite quality.
• HG/T 4967-2016 — Chinese chemical industry standard for pseudo-boehmite (used as Al source and binder).
• ASTM E1131-20 — Standard test method for compositional analysis by thermogravimetry. For TGA of coke burnoff profile.

Related Products & Resources

Aluminaworld supplies ZSM-5 and related catalyst materials for MTO, FCC additive, and xylene isomerization service:

Next Steps

If you are operating an MTO, FCC, xylene isomerization, toluene disproportion, ethylbenzene synthesis, or ammonia SCR unit on ZSM-5, and you want to extend catalyst life and reduce regeneration cost, our technical team can help. We provide:

• Free 200 g samples of fresh and 50-cycle regenerated ZSM-5 at your target Si/Al for side-by-side catalyst activity testing
• Full QC report with every shipment including XRD crystallinity, ICP-OES Si/Al, BET, NH3-TPD, and PSD
• Custom regeneration protocol design for your specific reactor and feed - we have supported DMTO, MTO swing, fixed-bed xylene isomerization, and FCC additive customers with tailored protocols
• 5 kg MOQ for R&D and regeneration trial, 500 kg for production orders
• Lead time 7 to 15 days from Zibo, Shandong
• Custom Si/Al grades (25 to 1000), custom particle sizes, custom metal impregnation (Pt, Pd, Cu, Fe, Ni, Co, Mo, Zn, Mg, P, La, Ce)

Contact our team via WhatsApp or email with your reactor configuration, current regeneration protocol, target cycle length, and annual ZSM-5 consumption. We will reply within one business day with a regeneration protocol recommendation and a quote for fresh or replacement catalyst.